Abstract
Background
A wide range of abnormalities has been reported in autistic brains, but these abnormalities may be the result of an earlier underlying developmental alteration that may no longer be evident by the time autism is diagnosed. The most consistent biological finding in autistic individuals has been their statistically elevated levels of 5hydroxytryptamine (5HT, serotonin) in blood platelets (platelet hyperserotonemia). The early developmental alteration of the autistic brain and the autistic platelet hyperserotonemia may be caused by the same biological factor expressed in the brain and outside the brain, respectively. Unlike the brain, blood platelets are shortlived and continue to be produced throughout the life span, suggesting that this factor may continue to operate outside the brain years after the brain is formed. The statistical distributions of the platelet 5HT levels in normal and autistic groups have characteristic features and may contain information about the nature of this yet unidentified factor.
Results
The identity of this factor was studied by using a novel, quantitative approach that was applied to published distributions of the platelet 5HT levels in normal and autistic groups. It was shown that the published data are consistent with the hypothesis that a factor that interferes with brain development in autism may also regulate the release of 5HT from gut enterochromaffin cells. Numerical analysis revealed that this factor may be nonfunctional in autistic individuals.
Conclusion
At least some biological factors, the abnormal function of which leads to the development of the autistic brain, may regulate the release of 5HT from the gut years after birth. If the present model is correct, it will allow future efforts to be focused on a limited number of gene candidates, some of which have not been suspected to be involved in autism (such as the 5HT_{4 }receptor gene) based on currently available clinical and experimental studies.
Background
Our ability to treat and prevent autism is severely limited by our lack of knowledge of what biological abnormality causes this developmental disorder. Since autism is considered primarily a brain disorder, much of the research over the past decades has focused on the autistic brain. Different groups have reported a wide range of anatomical abnormalities in autistic brains, such as reduced numbers of Purkinje cells in the cerebellum [13]; an unusually rapid growth of the cerebral cortical volume and head circumference during the first years after birth [49]; abnormal cortical minicolumns [1013]; abnormalities of the limbic system [1419]; abnormalities of the brainstem [2022]; and other brain alterations [2325].
Considering the complexity of brain development and its highly dynamic nature, these abnormalities may be the result of a long, complex chain of events. The original abnormality that caused them may occur early in development [26] and may be no longer obvious by the time autism is diagnosed. For example, an autisticlike loss of Purkinje cells may be caused by a mutation of the toppler gene, which causes severe ataxia in mice and appears to be irrelevant to autism [27]. Postmortem analysis of younger autistic brains is not an option, because it is usually not clear until age 2 or 3 which brains are autistic and which are not.
Fortunately, evidence suggests that at least one biological factor that causes the development of the autistic brain has a different function outside the central nervous system (CNS), where it continues to operate well into childhood and perhaps even into adulthood. Since the early 1960s, the most consistent biological finding in autistic individuals has been their statistically elevated serotonin (5hydroxytryptamine, 5HT) levels in blood platelets, or platelet hyperserotonemia [2833]. Unlike many of the reported alterations in the brain, this finding has been replicated numerous times by different groups, some of which have used large numbers of subjects. According to Anderson [33], "the platelet hyperserotonemia of autism [...] is generally considered to be one of the more robust and wellreplicated findings in biological psychiatry". The main reason why we have not capitalized on this major finding is that we have not been able to understand its origin or its relation to the brain.
It is unlikely that the autistic platelet hyperserotonemia is induced by the brain. The human bloodbrain barrier (BBB) becomes mature around one year after birth, if not earlier [34,35], and is virtually impenetrable to 5HT. Tryptophan, a 5HT precursor, can cross the BBB, but tryptophan levels do not appear to be altered in autistic individuals [36]. Unlike the anatomy of the mature brain, platelet 5HT levels should be actively maintained, because the halflife of platelets is only a few days [37,38]. This suggests that the factor that causes the platelet hyperserotonemia continues to be functionally active years after birth.
The statistical distribution of platelet 5HT levels in normal and autistic groups has certain characteristic features [31], but only recent studies have attempted to describe them in detail [39,40]. These distributions are likely to contain information about the underlying processes controlling platelet 5HT levels and, therefore, may help identify the factor that causes the platelet hyperserotonemia of autism. This same biological factor may be active during brain development (not necessarily in the same role), but there its identity may be obscured by the final complexity of a severalyearold autistic brain (Fig. 1). In the present study, published distributions of blood 5HT levels are analyzed by a novel, quantitative approach that may help trace early, experimentally undetectable brain abnormalities leading to autism.
Figure 1. A biological factor that causes autism may have a dual function. A factor that causes autism (shown in red) may be expressed (1) in the CNS, where it plays a role in the early development of the brain, and (2) outside the CNS, where it participates in processes that determine the 5HT levels in blood platelets. The "central" and "peripheral" 5HT systems are separated by the bloodbrain barrier (BBB) that matures after birth. It is usually not clear until age 2 or 3 whether the brain is autistic (black box). By that time, the factor has altered numerous developmental processes in the brain and may no longer be obvious. This same factor continues to operate years after birth outside the CNS, where it maintains higher than normal 5HT levels in blood platelets. In contrast to the brain, blood platelets are shortlived and continue to be produced throughout the life span.
Results
Basic model
The origin of the platelet hyperserotonemia of autism cannot be understood unless a certain model of the underlying physiological processes is accepted – whether it is an implicit model that is not clearly stated, a model described in words, or a mathematical model. One advantage of mathematical modeling is that it requires a clear description of all relevant interactions among the components of the system. Its greatest disadvantage is that sometimes clearcut choices have to be made where experimental data may suggest a few possible alternatives. In this section I introduce a model that is based on what is known about the 5HT circulation outside the CNS and point out two important but unresolved problems.
In search of a factor that can both cause platelet hyperserotonemia and alter normal brain function, many recent studies have focused on the serotonin transporter (SERT) that is expressed in blood platelets and brain neurons [41]. Despite early promising results [42], different groups have found little or no linkage [43] between SERT polymorphisms and autism in various ethnic groups [40,4447]. I have recently proposed [48] that the factor that interferes with brain development in autism may also regulate the release of 5HT from gut enterochromaffin (EC) cells, the main source of blood 5HT [36,49,50]. First, this hypothesis assumes that EC cells can monitor (directly or by way of gastrointestinal neurons) the 5HT levels in the surrounding extracellular space and can decrease or increase their 5HT release accordingly. Similar control mechanisms have long been suspected in the brain, where serotonergic neurons express 5HT autoreceptors [51,52]. Second, the levels of extracellular 5HT in the gut wall are assumed to be at equilibrium with the levels of free 5HT in the arterial blood. While the baseline extracellular levels of 5HT in the gut wall have not been precisely measured, the estimated levels of free 5HT in the arterial blood appear to be comparable to the extracellular 5HT levels in the brain [51,53], which expresses some of the same 5HT receptors as the gut [51,5457].
This hypothesis can be cast in a mathematical form. Suppose that EC cells indirectly monitor the levels of free 5HT that arrives in the gut with the arterial blood, compare these levels with the expected 5HT levels, and adjust their 5HT release to a new value (R_{n+1}), using a preset release value (R_{C}) as the reference point. The strength (gain) of this adjustment is controlled by a factor α, which is hypothesized to be different in normal and autistic individuals. After the blood leaves the gut, a large proportion (γ) of the free 5HT is quickly removed by the liver, lungs and other organs that express SERT and monoamine oxidases (MAOs) [5862]. The numerical value of γ is likely to vary from individual to individual, because the SERT and MAO genes have a number of polymorphic variants distributed in the population [40,45,46,6366]. Therefore, γ is considered to be a random variable with a known probability distribution. The model can then be described by the following system of equations:
F_{n + 1 }= (1  γ)F_{n }+ R_{n + 1}, (2)
Where (1  γ)F_{n }is the flux of free 5HT that enters the gut with the arterial blood, F_{C }is the preset ("expected") flux, and F_{n + 1 }is the flux of free 5HT that exits the gut (α ≥ 0, 0 ≤ γ ≤ 1, F_{C }> 0, R_{C }> 0). In the model, the 5HT release from EC cells does not include the 5HT that is used for local signaling and is rapidly removed by local gastrointestinal epithelial and neural cells expressing SERT [54,67,68]. This 5HT could be included in the model, together with the local clearance rate, if estimates of these parameters were available.
It is thought that little free 5HT is taken up by blood platelets, before most of it is removed by the liver, lungs and other organs [53,60]. Also, it has been suggested that platelet 5HT levels may depend on the levels of free 5HT in the blood almost linearly [53]. Then, at the steady state, F_{n + 1 }= F_{n }≡ F and R_{n + 1 }= R_{n }≡ R for any n, and platelet 5HT levels are
where K > 0 is a constant.
Note that ser(α, γ) is a decreasing function of γ. Also, at the steady state,
R = γF. (4)
It should be emphasized that the mathematical simplicity of equations (1) and (2) in no way implies that the biological regulation of 5HT release in the gut is simple. The human gut is a remarkably complex organ that uses a wide range of neurotransmitters and that may have at least as many neurons as the spinal cord [50]. Nevertheless, recent studies suggest that complex biological systems, such as brain neurons, can be "actively linear" [69], meaning that sophisticated biological mechanisms may act on intrinsically nonlinear physical processes to produce quantitative relationships that are mathematically linear.
The dependence of platelet 5HT levels on α and γ is plotted in Figure 2, where the numerical values of F_{C }and R_{C }are taken from previously published experimental and theoretical studies [48,53,70], and where the regulation of the 5HT release from EC cells is assumed to be less than fully functional in autistic individuals (note the low α value). A key feature of this dependence is that, in normal individuals, platelet 5HT levels remain low with any γ, whereas in autistic individuals these levels may be normal or higher than normal depending on the individual's γ. This dependence captures one of the most puzzling properties of the autistic distribution of platelet 5HT levels, which always overlaps with the control (normal) distribution, but always includes individuals whose 5HT levels are higher than normal [31]. It may also explain why the SERT and MAO genes may appear to be linked with autism but may not actually cause it. As shown in Figure 2, a low γ is a necessary but not sufficient condition for the platelet hyperserotonemia to occur. Given a low γ, the platelet hyperserotonemia will occur only in those individuals whose regulation of the 5HT release from EC cells is compromised (i.e., they are autistic and have a low α). It follows then that γ acts only as a modifier of platelet 5HT levels, and that the statistical distribution of γ may be the same in normal and autistic populations. Assuming an individual's γ value is determined, at least in part, by his/her variants of the SERT and MAO genes expressed in the liver, lungs and other organs, normal and autistic populations may have similar distributions of SERT and MAO polymorphisms. This assumption is supported by recent studies [40,4547,63,64].
Figure 2. Platelet levels as a function of α and γ. Platelet 5HT levels, ser(α, γ), plotted as a function of α (the factor regulating 5HT release from EC cells) and γ (the rate of 5HT clearance by the liver, lungs, and other organs). This relationship is described by equation (3), where K is a constant. Note that if α is normal (high), platelet 5HT levels stay low with any γ, but if α is autistic (low), individuals with a low γ become hyperserotonemic. The black circles mark the points whose coordinates are independent of α and are γ* = R_{C}/(R_{C }+ F_{C}) and ser(α, γ*) = KF_{C}. Note in equations (1) and (2) that R = R_{C }if and only if γ = γ*, so the distribution of γ is likely to contain γ*. This guarantees that the distributions of the 5HT levels in normal autistic groups will always overlap, as observed in clinical studies. For illustrative purposes, the normal and autistic values of α were arbitrarily set at 0.20 and 0.02, respectively. These are realistic values, as follows in the text. The other parameter values were taken from published studies [48, 53, 70] and were F_{C }= 210 ng/min and R_{C }= 3000 ng/min.
Two potentially contentious decisions were made in the model. First, the exact levels of free 5HT in the blood remain a debated issue. While a number of studies have found "low" but consistently measurable levels of free 5HT in the human blood [53,70,71], Chen et al. [72] have suggested that the concentration of free 5HT in the blood may be negligible, since these researchers have detected virtually no 5HT in the whole blood of SERTdeficient mice whose blood platelets cannot take up 5HT. Second, the model assumes that virtually all of the 5HT stored in blood platelets is taken up by them after the lungs, liver, and other organs have cleared a large proportion of the 5HT released by the gut. While evidence exists this may be the case [53,60], not all researchers agree. One could conceivably take into account both of these views by setting
ser(α, γ) ≡ K_{1 }F + K_{2}(1  γ)F
or, in a more general form,
where K_{1}, K_{2 }≥ 0 are constants and K(ω) is a function. However, this would require more detailed information about the dynamics of the 5HT uptake by platelets, which is not currently available [31].
Distributions generated by the model
While the model (Fig. 2) appears to capture some of the key characteristics of the reported platelet 5HT levels, it remains unclear whether it would produce similar results if α and γ took on other numerical values. The regulation of the 5HT release in EC cells is poorly understood and no experimental estimates for the parameter α are available. Is it actually lower in autistic individuals? Likewise, how reasonable is it to suppose that the distribution of γ is the same in normal and autistic groups? Importantly, would the model produce consistent numerical values of parameters if different experimental studies were used?
To answer these questions, one may consider the basic framework of the model to be correct, but make no a priori assumptions about the values of the parameters (with the exception of those that are experimentally known) or about their differences in normal and autistic individuals. Then the unknown parameters of the model may be allowed to vary in the numerical space until the statistical distributions of 5HT levels produced by the model closely match those reported in actual clinical studies. In order to be able to do this, one first has to find the theoretical statistical distributions of platelet 5HT levels produced by the model.
The exact population distribution of γ is unknown, but its mean value is likely to be close to one [60]. Since SERT gene polymorphisms may occur with comparable frequencies [73], the statistical distribution of γ in a population can be approximated by a continuous uniform distribution on the interval [a, b] with the probability density function
It can be shown from equations (3) and (5) that the probability density function of platelet 5HT levels then is
The theoretical population mean μ_{ser}(α, a, b) and variance (α, a, b) of platelet 5HT levels follow immediately:
and
where U ≡ F_{C } R_{C}α.
The standard deviation of platelet 5HT levels in the population then is
Distributions reported in clinical studies
Mean values of normal and autistic blood 5HT levels have been reported and discussed in numerous publications [2833]. In contrast, the precise statistical distributions of the platelet 5HT levels in normal and autistic groups, such as their histograms (which roughly approximate their theoretical probability density functions), have so far attracted little attention. Only a few recent reports have presented more detail about the shape of these distributions. These reports are used in the following analysis:
(i) Mulder et al. [39] is recent and perhaps the most reliable report to date. It has used a relatively large sample of subjects whose platelet 5HT levels are presented in histograms. The authors of this report are wellestablished researchers of blood 5HT and autism. One of the coauthors, G.M. Anderson, has had numerous publications on the subject over the past several decades.
(ii) Coutinho et al. [40] have studied a large sample of subjects and presented their 5HT levels in histograms, also explicitly listing their minimum and maximum values. However, their reported mean 5HT levels are somewhat low, and the autistic 5HT levels are higher than, but not significantly different from, the normal 5HT levels.
(iii) McBride et al. [74] is a detailed report on the means and standard variations of platelet 5HT levels in ethnically different groups, but the data are not presented in histogram form. Here, the minimum and maximum values of the distributions are recovered from their Figure 1, and the pooled means of the prepubertal children are recalculated from their Table 2.
Table 2. Predicted and observed ranges, means (<ser>), and standard deviations (SD) of platelet 5HT levels, ser(α, γ). The distribution of γ was assumed to be continuously uniform; the theoretical SD values given in the table can be further improved by assuming that γ has a beta distribution or a normal distribution (see the text). Note that, strictly speaking, the model's <ser > and SD are precise theoretical expectations and standard deviations and, therefore, the notation μ_{ser }(α, a, b) and σ_{ser }(α, a, b) would be more accurate (but less convenient here).
It is important to note that these reports are the only ones presently available and, therefore, no selection bias was introduced by choosing them for the present study.
Finding α and [a, b] from clinical data
In order to be able to compare the model's predictions with actual clinical reports, the numerical output of the model has to be scaled to the units of the used experimental studies. This scaling can be done by adjusting the parameter K in equation (3). The studies have reported the following means of the blood 5HT levels in their normal groups: 3.58 nmol/10^{9 }platelets [39], 260 ng/10^{9 }platelets [40], and 230 ng/ml [74]. The last number was obtained by pooling the reported prepubertal means of the three ethnic groups. Assuming the flux of free 5HT to the gut is around 210 ng/min in normal individuals [48,53,70], it follows from equation (3) that
where <...> denotes experimentally obtained means. Now we can calculate the approximate K values for each of the studies by dividing their reported mean 5HT levels by the approximate flux of free 5HT to the gut. This yields the following K values for the reports of Mulder et al. [39], Coutinho et al. [40] and McBride et al. [74], respectively: 0.0170 (nmol min ng^{1 }10^{9 }platelets), 1.2381 (min 10^{9 }platelets), and 1.0952 (min ml^{1}).
Next, we try to find such numerical values of [a, b], α_{normal}, and α_{autistic}, that they minimize the difference between the predicted and observed levels of blood 5HT. Suppose that the observed levels of blood 5HT vary from Min^{OBS }to Max^{OBS }and that the observed mean of blood 5HT is <ser>^{OBS}. The following error function can then be constructed:
where
and i = normal, autistic.
Note that, compared with the mismatch between the predicted and observed ranges of the distributions, the mismatch between the predicted and observed means is penalized "twice as much", because observed means are likely to be more accurate than observed minimal and maximal values.
This error function was numerically minimized by using the standard NelderMead (downhill simplex) and differential evolution methods [75] implemented in Mathematica's NMinimize function (Wolfram Research, Inc.). Since the values of R_{C }and F_{C }may be approximated from published studies but are not necessarily accurate, R_{C }was centered at 3000 ng/min based on a published estimate [53] and was allowed to vary ± 33%, whereas the value of F_{C }was centered at 210 ng/min based on published estimates [48,53,70] and was allowed to vary ± 50% (more variation was allowed for F_{C }because less is known about its actual value). No constraints were set for the interval [a, b] (i.e., 0 ≤ a <b ≤ 1). The variables α_{normal }and α_{autistic }were allowed to vary from 0 to 5 and no a priori assumptions were made about their relative values (i.e., both α_{normal }>α_{autistic }and α_{normal }≤ α_{autistic }were allowed). It can be shown that the system (equations (1) and (2)) is stable if 0≤α<F_{C}(2  γ)/[R_{C}(1  γ)]. Since the system should be stable for any γ ∈[a, b] and [a, b] is likely to contain the point γ ≈ 0.99 [60] or γ ≈ 0.93 [48], choosing α between 0 and 5 allows the optimization procedure to use virtually any value of α where the system maintains stability.
The numerical values of the model's parameters (α_{normal}, α_{autistic}, [a, b], F_{C}, and R_{C}) that minimized the error function are given in Table 1. Note that all three clinical studies yielded similar sets of values. Most importantly, the minimization algorithms yielded the best match between the model and the clinical reports when α_{autistic }was virtually zero.
Table 1. Estimates of F_{C}, R_{C}, a, b, α_{normal}, and α_{autistic}, obtained by numerical minimization of the error function.
By plugging these obtained values of the parameters into equations (12), (13), (14) and (9), one can obtain the values of 5HT levels predicted by the model and compare them with the actual observed levels. As shown in Table 2, the predicted values closely match the values observed in Mulder et al. [39] and McBride et al. [74]. The largest mismatch was between the predicted and observed minimal values. The model predicted slightly higher mean 5HT levels for Coutinho et al. [40] than were actually observed; interestingly, Coutinho et al. [40] have in fact reported unusually low platelet 5HT levels.
Distribution of γ can be approximated by beta and normal distributions
One advantage of choosing the uniform distribution to represent γ is that it simplifies calculations and allows finding the exact formulae for means and standard deviations. However, the model tends to overestimate the standard deviations of platelet 5HT levels (Table 2), because in the uniform distribution even extreme γ values occur with same probability as all others. Instead of approximating the distribution of γ as uniform, one may want a distribution of which the probability density function drops off more smoothly near the minimal and maximal values. This can be achieved by replacing the uniform distribution of γ with the beta distribution, the uniform distribution being its special case [76]. The following deals with mathematical technicalities of this replacement. Nonmathematically inclined readers may skip them and go immediately to Figures 4 and 5 referred to at the end of this section.
Figure 4. Model replicates published data. A, B, The model's simulation of Mulder et al.'s sampling [39], assuming γ has the beta distribution on the interval [0.8060, 0.9612] with both shape parameters equal to 1.5. The platelet 5HT levels were calculated by using equation (3), with the values of K, F_{C}, R_{C}, α_{normal }and α_{autistic }taken from Table 1. C, D, The actual data from Mulder et al. [39] (reprinted by permission from Lippincott Williams & Wilkins, modified). In the simulated and actual sampling, 60 normal and 33 autistic subjects were used. Note that the exact appearance of the histograms will vary from sampling to sampling due to the small number of cases in each bin.
Figure 5. Model predicts the shape of the normal and autistic distributions of platelet 5HT levels. Histograms obtained by simulating a sampling of a very large number of normal and autistic individuals (a million subjects in each group). The distribution of γ was assumed to be (A, B) the beta distribution on the interval [0.8060; 0.9612] with both shape parameters equal to 1.5 (see the text); or (C, D) the normal (Gaussian) distribution with mean 0.8836 (the midpoint of the interval [0.8060; 0.9612]) and standard deviation 0.04 (see the text). The platelet 5HT levels were calculated by using equation (3), with the values of K, F_{C}, R_{C}, α_{normal }and α_{autistic }taken from Table 1. In a very large sampling, the number of cases in each histogram bin closely approximates the number of cases predicted by the exact probability distribution functions. The Chisquare test confirmed that the normal and autistic distributions predicted by the model may underlie the distributions reported by Mulder et al. (2004). The following goodnessoffit results were obtained: = 12.38 (P = 0.26) and = 11.29 (P = 0.19) for the normal and autistic groups, respectively, if γ had the beta distribution; and = 13.36 (P = 0.27) and = 12.21 (P = 0.14) for the normal and autistic groups, respectively, if γ had the normal distribution (bins were pooled if theoretical bins had fewer than 3 cases). It is important that both the normal and autistic distributions had the same underlying distribution of γ and that only one parameter, α, was needed to switch from the normal distribution to the autistic distribution. Also, compare the histograms in C and D, based on the data of Mulder et al. [39], with those in Figure 1 of Coutinho et al. [40].
Note that if the obtained parameter values (Table 1) are plugged into equation (3), the normal and autistic platelet 5HT levels turn out to depend on γ almost linearly (Fig. 3). This allows "warping" the uniform distribution of γ into a symmetric beta distribution on the same interval, with little effect on the theoretical mean values of ser(α, γ). Suppose that γ has a symmetric beta distribution on [a, b], whose shape is determined by the parameters m and n, such that m = n (if m = n = 1, the beta distribution becomes the uniform distribution). We can use a Taylor series to formally linearize ser(α, γ) around γ_{0 }= (a + b)/2 as ser(α, γ) ≈ ser(α, γ_{0})  λ (γ  γ_{0}) ≡ serL(α, γ),
Figure 3. Platelet levels plotted with the parameter values derived from published studies. Platelet 5HT levels as functions of γ for the data of Mulder et al. [39], Coutinho et al. [40] and McBride et al. [74]. Equation (3) and the estimated parameter values from Table 1 were used. The arrowheads mark the predicted intervals of the γ distributions (Table 1). For comparison, the Yaxes were scaled proportionally to the K values of the three studies (Table 1).
Then, keeping in mind that γ has a beta distribution, the standard deviation of serL(α, γ) becomes
Since the values of λ, a, and b have already been estimated (Table 1), it is now possible to obtain the m values that yield such standard deviations of the linearized ser(α, γ) that they precisely match those reported in the clinical studies (Table 2). The following m values were obtained for the normal and autistic groups, respectively: 1.2940 and 1.7028 for the data of Mulder et al. [39]; and 1.8308 and 1.8748 for the data of Coutinho et al. [40]. Pooled standard variations were unavailable in McBride et al. [74]. We have earlier assumed that normal and autistic groups have the same γ distribution. Therefore, the actual m values can be approximated by 1.50 for Mulder et al. [39] and 1.85 for Coutinho et al. [40].
Likewise, γ can be assumed to have a normal distribution with mean (a + b)/2 and standard deviation σ. Then the standard deviation of serL(α, γ) becomes
σ_{serL}(α, a, b, σ) = λσ, (17)
where λ is the same as in equation (15), and we obtain the following σ values for the normal and autistic groups, respectively: 0.0410 and 0.0370 for the data of Mulder et al. [39]; and 0.0630 and 0.0624 for the data of Coutinho et al. [40]. Therefore the actual σ values can be approximated by 0.04 for Mulder et al. [39] and 0.06 for Coutinho et al. [40].
The model now easily generates "normal" and "autistic" samples of platelet 5HT levels that closely match the actual reported data (Fig. 4). Most importantly, the switch from the normal distribution to the autistic distribution requires changing only one parameter, α.
It is not known what normal and autistic distributions would look like if one could sample a very large number of subjects. The model can predict the shape of these distributions by simulating such large sampling (Fig. 5).
Is the 5HT synthesis rate altered in autism?
One of the most important questions in autism research is whether the rate of 5HT synthesis is altered in the brain and gut of autistic individuals. If 5HT synthesis is altered in the autistic brain, as some studies have suggested [7779], this potentially may have a great impact on brain development [80,81] (but caution should be exercised in predicting the extent of these alterations [82]).
The brain 5HT and the gut 5HT are synthesized by two different tryptophan hydroxylases [49] that, at least in humans, have different properties and are regulated differently [83]. While the biological factor underlying the parameter α of the model is hypothesized to play a role in the developing brain (Fig. 1), the model makes no assumptions about its exact function in the brain. In the brain, it may not regulate 5HT release from serotonergic neurons and may have a different function (see, for example, Figure 4 of [48]). Therefore, this section focuses only on the 5HT synthesis and release in the gut.
It is important to note that the model says nothing about the rate of 5HT synthesis in the gut and rather deals with the rate of 5HT release from the gut. However, most clinical and experimental studies make no such distinction and, therefore, their relevance to the model is discussed assuming higher 5HT synthesis rates do lead to higher 5HT release rates.
It follows from equations (3) and (4) that, at the steady state,
and that this relationship is independent of γ. This means that if one were to sample any group of individuals and could measure their platelet 5HT levels and gut 5HT release rates precisely, the correlation coefficient between these two variables would always be minus one, irrespective of the distribution of γ. In other words, equation (18) predicts that individuals with higher platelet 5HT levels should have lower 5HT release rates.
How can lower 5HT release rates lead to higher platelet 5HT levels? Note that, in the model, both the platelet 5HT levels and the 5HT release rate are dynamically linked through the 5HT clearance rate, γ. As γ grows lower, less 5HT is removed from the system and more of 5HT is accumulated in blood platelets. At the same time, these higher 5HT levels drive down the 5HT release rate in the gut, as required by equation (1).
Still, it appears that the results of clinical studies are inconsistent with equation (18). Three important findings should be noted:
(i) Minderaa et al. [36] have found no significant correlation between whole blood 5HT levels and 5HT synthesis in the gut, measured as the production of urinary 5HIAA [36]. Similar results have been obtained by Launay et al. [84] and other groups (reviewed in [31]).
(ii) Croonenberghs et al. [85] have shown that the 5HT synthesis in the gut of autistic individuals may be higher than that in normal individuals, at least when subjects are administered 5hydroxytryptophan (5HTP), an immediate precursor of 5HT.
(iii) Carcinoid tumors, derived from gut EC cells, may result in excessive synthesis and release of 5HT, which in turn may lead to elevated platelet 5HT levels [86].
A more careful analysis reveals that these findings are not only consistent with the model, but that the model can reconcile some of the apparent contradictions among them:
(i) It follows from the model that the measured correlation between platelet 5HT levels and 5HT release rates should be close to zero in autistic groups, even though equation (18) holds.
In fact, we can rewrite equation (18) as
Now consider two random variables, η and ξ, that are linearly dependent such that
η = wξ + q, (20)
where w and q are constants. It follows from equation (20) that the correlation between them is either 1 or 1, depending on the sign of w.
Denote the means of these variables μ_{η }and μ_{ξ}, respectively, and their standard deviations σ_{η }and σ_{ξ}, respectively. Suppose next that the errors of measurement of η and ξ are independent random variables ε_{η }and ε_{ξ}, such that their expected values are zero and standard deviations are δ_{η }and δ_{ξ}, respectively. Note that experimentally we can measure only η* = η + ε_{η }and ξ* = ξ + ε_{ξ}. The expected values of η* and ξ* are the same as those of η and ξ. However, the theoretical correlation coefficient between η* and ξ* now becomes
If the standard deviations of the errors of measurement are small, we obtain ρ(η*, ξ*) ≈ ± 1, as expected from equation (20).
Now we return to equation (19). Any experimental measurement of R (5HT release) and ser(α, γ) (platelet 5HT levels) will contain a measurement error. Denoting these measured values ser*(α, γ) and R*, one obtains from equations (19), (20), and (21) that the correlation coefficient between R* and ser*(α, γ) is
where
w = (αR_{C})/(KF_{C}), (23)
σ_{ser }> 0 is the standard deviation of ser(α, γ), and δ_{R} > 0 and δ_{ser} > 0 are the standard deviations of the errors of measurement of R and ser(α, γ), respectively. The estimated values of K, F_{C}, R_{C}, and α can be obtained from Table 1 and the values of σ_{ser }from Table 2 or from the original published data.
Consider now an autistic group whose α → 0 (Table 1). Then, from equation (23), w → 0, and it follows from equation (22) that .
(ii) Croonenberghs et al. [85] have recently shown that oral administration of 5hydroxytryptophan (5HTP) leads to higher platelet 5HT levels in autistic patients, and the authors have suggested that the 5HT synthesis rate may be higher in the gut of autistic subjects compared with normal subjects.
Suppose that the administered 5HTP is converted to 5HT at the same rate in both normal and autistic groups. It is likely that the exogenous influx of 5HTP results in a comparable exogenous influx of 5HT, because the ratelimiting step in the synthesis of 5HT is not the 5HTP conversion to 5HT, but rather the tryptophan conversion to 5HTP [87].
Notice that the system is not in its steady state during the experiment and, therefore, we have to use equations (1) and (2), which now should contain the exogenous source of 5HT. It is straightforward to see that the system then becomes
F_{n + 1 }= (1  γ)F_{n }+ R_{n + 1 }+ R_{EX}, (25)
where R_{EX }is the exogenous flux of 5HT.
Solving equations (24) and (25) stepbystep essentially replicates the major finding of Croonenberghs et al. [85] (Fig. 6). However, the model predicts that the higher blood 5HT levels in autistic subjects are not due to a higher 5HT synthesis rate, but rather to the failure of their gut to decrease the release of endogenous 5HT in response to the high concentration of 5HT caused by the administration of 5HTP.
Figure 6. An exogenous source of 5HT elevates platelet 5HT levels in an autistic group more than in a normal group. For the simulation, the initial values of platelet 5HT levels (ser(α, γ)) and 5HT release rate (R) were set at zero and the system developed according to equations (1) and (2). After the system reached its steady state, an exogenous 5HT source was "turned on" (+R_{EX}) and the system developed according to equations (24) and (25). After 5 steps, the exogenous 5HT source was "turned off" (R_{EX}) and the system developed according to equations (1) and (2) until it returned to its steady state. Each point is the mean of 10,000 simulated individuals whose γ had the beta distribution on the interval [0.8060, 0.9612] (see Table 1) with both shape parameters equal to 1.5 (see the text). Individual plots (not shown) looked essentially the same as the mean plots. The ratio between the autistic and normal platelet 5HT levels (A) at step 7 (at the steady state) is 1.25 and the same ratio at step 13 is 1.35. The numerical values of the parameters were K = 0.0170, F_{C }= 105, R_{C }= 2000, α_{normal }= 0.1510, α_{autistic }= 0.0000 (Table 1) and R_{EX }= 4000. Compare these plots with Figure 1 of Croonenberghs et al. [85].
(iii) In the case of carcinoid tumors, abnormally large amounts of 5HT may be released into the blood. It is likely that the normal mechanisms regulating 5HT release are compromised or absent in carcinoid tumors. Then instead of equations (1) and (2) one can consider only one equation (2), which can be rewritten as
F_{n + 1 }= (1  γ)F_{n }+ R_{CARCINOID}, (26)
where R_{CARCINOID }is large and relatively constant. Then, at the steady state,
F = R_{CARCINOID}/γ
and
It is obvious that in this abnormal case higher 5HT release rates will lead to higher platelet 5HT levels, as reported by Kema et al. [86].
Discussion
The presented model is based on the hypothesis that at least one factor that interferes with normal brain development in autism also participates in the regulation of 5HT release from enterochromaffin cells. When applied to the data of three published studies, the model predicts that this factor is virtually nonfunctional in autistic individuals (Table 1).
Before the biological nature of this factor is discussed, it should be noted that the parameter values obtained for each of the three published studies were virtually the same (Table 1). This underlying consistency of the data is not trivial, since Mulder et al. [39] have suggested that their autistic distribution may be bimodal and thus qualitatively different from the control (normal) distribution, whereas Coutinho et al. [40] have reported a clearly unimodal autistic distribution that so overlapped with the control distribution that their means were not statistically significant. It should also be noted that initially γ was allowed to vary from zero to one, but the numerical optimization based on the published data narrowed this range down to approximately 0.8 – 1.0 (Table 1). This agrees well with actual experimental data. An early study has approximated the dog's γ as 0.99 and shown that the 5HT clearance by the lungs varies from 0.80 to 0.98 [60]. The mean human γ may be somewhat smaller, because the rate of 5HT release by gut enterochromaffin cells has been predicted to be around 3000 ng/min [53] and the arterial flow of free 5HT has been estimated to be around 210 ng/min [48,53,70]. This suggests that, in humans, approximately 93% of free 5HT is cleared in one circulation and, therefore, the value of γ is close to 0.93. The model predicted similar γ distributions in normal and autistic groups, supporting the hypothesis that the frequencies of SERT and MAO polymorphisms in normal and autistic groups may be the same.
The most significant result is that the factor that regulates 5HT release from EC cells (represented by the parameter α) appears to be virtually nonfunctional in autistic individuals (Table 1). What is the biological nature of α? Evidence suggests that EC cells may express 5HT_{3}, 5HT_{4 }and 5HT_{1A }receptors [55,8890] and that they may also express 5HT_{2 }receptors [89]. Some of these receptors appear to be involved in the autoregulation of 5HT release [89,90]. While one report has failed to find 5HT_{3 }and 5HT_{4 }receptor mRNAs in cultured EC cells [91], the regulation of 5HT release from EC cells may also be indirect, by way of enteric neurons. These neurons are known to express various 5HT receptors [54,55,92,93] and can control 5HT release from EC cells by acting on their cholinergic and other receptors [88,9496].
The model is based on a negative feedback loop. It has been shown that such negative feedback may be mediated by 5HT_{4 }receptors expressed by EC cells and that this negative feedback appears to dominate over the positive feedback mediated by 5HT_{3 }receptors [89,90]. A recent study has suggested that under normal circumstances (as opposed to conditions such as carcinoid tumors) the concentration of endogenous 5HT may not be high enough to activate 5HT_{4 }receptors and alter the 5HT release from EC cells [89]. At least superficially, this mirrors recent findings in the brain, where 5HT_{1A }and 5HT_{1B }receptors, long assumed to act as autoreceptors, may not actually be activated by extracellular 5HT unless its concentration reaches excessive levels [51]. Since precise measurements of 5HT release in the gut and the brain are difficult, it is more likely that these receptors do control 5HT release under normal circumstances, but that their effect on 5HT release is more subtle than we expect. The model's small value of α appears to predict such subtle regulation.
Can 5HT_{4 }receptors be involved in autism? One agonist used to study the effects of 5HT_{4 }receptors on the 5HT release from EC cells has been 5methoxytryptamine (5MT) [89,90], which has high affinity for these receptors [57]. While 5MT has been reported to inhibit the 5HT release from EC cells, subcutaneous 5MT injections in pregnant rats produces pups with autisticlike symptoms [97] and subcutaneous 5MT injections in pregnant mice may lead to an autisticlike disruption of cortical columns in the pups [11,81]. Normal brain development may be altered if brain 5HT_{4 }receptors are compromised, because these receptors appear to be expressed in the marginal zone of the adult human brain [98] and, therefore, may also be expressed in CajalRetzius cells of the developing brain. It has been recently shown that an abnormal serotonergic input to CajalRetzius cells during development may lead to autisticlike cortical abnormalities [81]. Interestingly, the expression of the 5HT_{4 }receptor is very low in the cerebral cortex of the guinea pig [99], suggesting that this receptor may play a specific role in the primate brain. Generally, we are only beginning to understand the role of the 5HT_{4 }in brain development, because the human 5HT_{4 }receptor gene consists of at least 38 exons and at least eight Cterminal splice variants of the human 5HT_{4 }receptor have been described [57].
Other 5HT receptors, as well as other mechanisms, may be involved both in the regulation of 5HT release from the gut and in brain development. For example, 5HT_{1A }and 5HT_{2 }receptors have been implicated in autism [31,100102]. As already discussed, these receptors can also regulate the 5HT release from EC cells. Moreover, 5MT is a rather nonspecific 5HT receptor agonist [103] and appears to be colocalized with 5HT in most brain neurons [104]. Therefore, some of its effects may be produced by its acting on a few types of 5HT receptors at the same time, both in the gut and the brain.
The model assumes that the 5HT clearance rate (γ) and the gain of 5HT release (α) are independent. Generally, the expression of neurotransmitter receptors or their sensitivity can dynamically change depending on the availability of the neurotransmitter. For example, gut 5HT_{3 }receptors undergo structural and functional changes in SERTknockout mice [105] and 5HT_{1A }receptors in the human brain have different affinities in individuals with different SERT polymorphic variants [106]. These and other related findings are likely to become indispensable for understanding the platelet hyperserotonemia of autism; unfortunately, too little information is currently available for quantitative modeling of these relationships.
Intriguingly, α may be represented by biological mechanisms other than 5HT receptors. For example, adenosine and ATP may modulate the 5HT release from human EC cells [107,108] and ATP also activates microglia in the brain [109]. A study, called by some researchers "the most important postmortem study of autism to date" [110], has found an abnormal activation of microglia in autistic brains [111].
It should be noted in conclusion that the mathematical framework of the model allows it to be modified so that it no longer depends on free 5HT in the blood. In fact, one could conceivably build a model where 5HT is released by EC cells, cleared by SERTexpressing cells locally, and where the remaining extracellular 5HT acts on the mechanisms controlling 5HT release from EC cells, without leaving the gut. Assuming γ now denotes the local clearance and α is the gain of the 5HT release, one again may arrive at a system of equations similar to equations (1) and (2).
Conclusion
The origin of autism is as much a conceptual problem as it is experimental. The theoretical approach introduced here brings together information on the "central" and "peripheral" 5HT and offers new insights into early abnormalities of the developing autistic brain that may otherwise escape direct experimental detection.
Methods
All symbolic and numerical calculations were done in Mathematica 5.0.0, 5.0.1, 5.1.0, or 5.1.1 (Wolfram Research, Inc.). Where the numerical minimization of the error function produced different sets of numerical values in different releases of Mathematica, the values that yielded the smallest error were used (for the purpose of this study, Mathematica 5.1.1 was superior to the earlier releases). The figures were generated in Mathematica and prepared for publication in Adobe Illustrator 10 or CS (Adobe Systems, Inc.).
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
SJ conceived of and carried out the presented study.
Acknowledgements
I thank Dr. P. Rakic and the National Alliance for Autism Research (NAAR) for their financial support, the anonymous reviewers for their valuable suggestions, and Dr. G.M. Anderson, Dr. A.E. Ayoub and Michael Fischer for their comments on the revised manuscript. I also thank Vaiva, my inspiration.
References

Williams RS, Hauser SL, Purpura DP, DeLong GR, Swisher CN: Autism and mental retardation: neuropathologic studies performed in four retarded persons with autistic behavior.
Arch Neurol 1980, 37:749753. PubMed Abstract

Ritvo ER, Freeman BJ, Scheibel AB, Duong T, Robinson H, Guthrie D, Ritvo A: Lower Purkinje cell counts in the cerebella of four autistic subjects: initial findings of the UCLANSAC Autopsy Research Report.
Am J Psychiatry 1986, 143:862866. PubMed Abstract

Kern JK: Purkinje cell vulnerability and autism: a possible etiological connection.
Brain Dev 2003, 25:377382. PubMed Abstract  Publisher Full Text

Woodhouse W, Bailey A, Rutter M, Bolton P, Baird G, Le CA: Head circumference in autism and other pervasive developmental disorders.
J Child Psychol Psychiatry 1996, 37:665671. PubMed Abstract

Courchesne E, Karns CM, Davis HR, Ziccardi R, Carper RA, Tigue ZD, Chisum HJ, Moses P, Pierce K, Lord C, Lincoln AJ, Pizzo S, Schreibman L, Haas RH, Akshoomoff NA, Courchesne RY: Unusual brain growth patterns in early life in patients with autistic disorder: an MRI study.
Neurology 2001, 57:245254. PubMed Abstract  Publisher Full Text

Aylward EH, Minshew NJ, Field K, Sparks BF, Singh N: Effects of age on brain volume and head circumference in autism.
Neurology 2002, 59:175183. PubMed Abstract  Publisher Full Text

Courchesne E, Carper R, Akshoomoff N: Evidence of brain overgrowth in the first year of life in autism.
JAMA 2003, 290:337344. PubMed Abstract  Publisher Full Text

Carper RA, Moses P, Tigue ZD, Courchesne E: Cerebral lobes in autism: early hyperplasia and abnormal age effects.
Neuroimage 2002, 16:10381051. PubMed Abstract  Publisher Full Text

Carper RA, Courchesne E: Inverse correlation between frontal lobe and cerebellum sizes in children with autism.
Brain 2000, 123 ( Pt 4):836844. PubMed Abstract  Publisher Full Text

Casanova MF, Buxhoeveden DP, Brown C: Clinical and macroscopic correlates of minicolumnar pathology in autism.
J Child Neurol 2002, 17:692695. PubMed Abstract  Publisher Full Text

Casanova MF, Buxhoeveden DP, Switala AE, Roy E: Minicolumnar pathology in autism.
Neurology 2002, 58:428432. PubMed Abstract  Publisher Full Text

Casanova MF, Buxhoeveden D, Gomez J: Disruption in the inhibitory architecture of the cell minicolumn: implications for autisim.
Neuroscientist 2003, 9:496507. PubMed Abstract  Publisher Full Text

Casanova MF: White matter volume increase and minicolumns in autism.
Ann Neurol 2004, 56:453. PubMed Abstract  Publisher Full Text

Raymond GV, Bauman ML, Kemper TL: Hippocampus in autism: a Golgi analysis.
Acta Neuropathol (Berl) 1996, 91:117119. PubMed Abstract  Publisher Full Text

Blatt GJ, Fitzgerald CM, Guptill JT, Booker AB, Kemper TL, Bauman ML: Density and distribution of hippocampal neurotransmitter receptors in autism: an autoradiographic study.
J Autism Dev Disord 2001, 31:537543. PubMed Abstract  Publisher Full Text

Kemper TL, Bauman ML: Neuropathology of infantile autism.
Mol Psychiatry 2002, 7 Suppl 2:S12S13. PubMed Abstract  Publisher Full Text

Amaral DG, Bauman MD, Schumann CM: The amygdala and autism: implications from nonhuman primate studies.
Genes Brain Behav 2003, 2:295302. PubMed Abstract  Publisher Full Text

Bauman M, Kemper TL: Histoanatomic observations of the brain in early infantile autism.
Neurology 1985, 35:866874. PubMed Abstract

Schumann CM, Hamstra J, GoodlinJones BL, Lotspeich LJ, Kwon H, Buonocore MH, Lammers CR, Reiss AL, Amaral DG: The amygdala is enlarged in children but not adolescents with autism; the hippocampus is enlarged at all ages.
J Neurosci 2004, 24:63926401. PubMed Abstract  Publisher Full Text

Rodier PM, Ingram JL, Tisdale B, Nelson S, Romano J: Embryological origin for autism: developmental anomalies of the cranial nerve motor nuclei.
J Comp Neurol 1996, 370:247261. PubMed Abstract  Publisher Full Text

Rodier PM: Converging evidence for brain stem injury in autism.
Dev Psychopathol 2002, 14:537557. PubMed Abstract  Publisher Full Text

Rodier PM: The early origins of autism.
Sci Am 2000, 282:5663. PubMed Abstract

Belmonte MK, Allen G, BeckelMitchener A, Boulanger LM, Carper RA, Webb SJ: Autism and abnormal development of brain connectivity.
J Neurosci 2004, 24:92289231. PubMed Abstract  Publisher Full Text

Palmen SJ, Van EH, Hof PR, Schmitz C: Neuropathological findings in autism.
Brain 2004. PubMed Abstract  Publisher Full Text

Courchesne E, Redcay E, Kennedy DP: The autistic brain: birth through adulthood.
Curr Opin Neurol 2004, 17:489496. PubMed Abstract  Publisher Full Text

Chugani DC: Serotonin in autism and pediatric epilepsies.
Ment Retard Dev Disabil Res Rev 2004, 10:112116. PubMed Abstract  Publisher Full Text

Duchala CS, Shick HE, Garcia J, Deweese DM, Sun X, Stewart VJ, Macklin WB: The toppler mouse: A novel mutant exhibiting loss of Purkinje cells.
J Comp Neurol 2004, 476:113129. PubMed Abstract  Publisher Full Text

Schain RJ, Freedman DX: Studies on 5hydroxyindole metabolism in autistic and other mentally retarded children.
J Pediatr 1961, 58:315320. PubMed Abstract

Hanley HG, Stahl SM, Freedman DX: Hyperserotonemia and amine metabolites in autistic and retarded children.
Arch Gen Psychiatry 1977, 34:521531. PubMed Abstract

Anderson GM, Freedman DX, Cohen DJ, Volkmar FR, Hoder EL, McPhedran P, Minderaa RB, Hansen CR, Young JG: Whole blood serotonin in autistic and normal subjects.
J Child Psychol Psychiatry 1987, 28:885900. PubMed Abstract

Anderson GM, Horne WC, Chatterjee D, Cohen DJ: The hyperserotonemia of autism.
Ann N Y Acad Sci 1990, 600:331340. PubMed Abstract

Cook EHJ: Brief report: pathophysiology of autism: neurochemistry.
J Autism Dev Disord 1996, 26:221225. PubMed Abstract  Publisher Full Text

Anderson GM: Genetics of childhood disorders: XLV. Autism, part 4: serotonin in autism.
J Am Acad Child Adolesc Psychiatry 2002, 41:15131516. PubMed Abstract  Publisher Full Text

Wenzel D, Felgenhauer K: The development of the bloodCSF barrier after birth.
Neuropadiatrie 1976, 7:175181. PubMed Abstract

Virgintino D, Errede M, Robertson D, Capobianco C, Girolamo F, Vimercati A, Bertossi M, Roncali L: Immunolocalization of tight junction proteins in the adult and developing human brain.
Histochem Cell Biol 2004, 122:5159. PubMed Abstract  Publisher Full Text

Minderaa RB, Anderson GM, Volkmar FR, Akkerhuis GW, Cohen DJ: Urinary 5hydroxyindoleacetic acid and whole blood serotonin and tryptophan in autistic and normal subjects.
Biol Psychiatry 1987, 22:933940. PubMed Abstract  Publisher Full Text

Baker GR, Sullam PM, Levin J: A simple, fluorescent method to internally label platelets suitable for physiological measurements.
Am J Hematol 1997, 56:1725. PubMed Abstract  Publisher Full Text

Heyssel RM: Determination of human platelet survival utilizing C14labeled serotonin.
J Clin Invest 1961, 40:21342142. PubMed Abstract  PubMed Central Full Text

Mulder EJ, Anderson GM, Kema IP, De BA, Van Lang ND, Den Boer JA, Minderaa RB: Platelet Serotonin Levels in Pervasive Developmental Disorders and Mental Retardation: Diagnostic Group Differences, WithinGroup Distribution, and Behavioral Correlates.
J Am Acad Child Adolesc Psychiatry 2004, 43:491499. PubMed Abstract  Publisher Full Text

Coutinho AM, Oliveira G, Morgadinho T, Fesel C, Macedo TR, Bento C, Marques C, Ataide A, Miguel T, Borges L, Vicente AM: Variants of the serotonin transporter gene (SLC6A4) significantly contribute to hyperserotonemia in autism.
Mol Psychiatry 2004, 9:264271. PubMed Abstract  Publisher Full Text

Lesch KP, Wolozin BL, Murphy DL, Reiderer P: Primary structure of the human platelet serotonin uptake site: identity with the brain serotonin transporter.
J Neurochem 1993, 60:23192322. PubMed Abstract

Cook EHJ, Courchesne R, Lord C, Cox NJ, Yan S, Lincoln A, Haas R, Courchesne E, Leventhal BL: Evidence of linkage between the serotonin transporter and autistic disorder.
Mol Psychiatry 1997, 2:247250. PubMed Abstract  Publisher Full Text

Spielman RS, Ewens WJ: The TDT and other familybased tests for linkage disequilibrium and association.
Am J Hum Genet 1996, 59:983989. PubMed Abstract

McCauley JL, Olson LM, Dowd M, Amin T, Steele A, Blakely RD, Folstein SE, Haines JL, Sutcliffe JS: Linkage and association analysis at the serotonin transporter (SLC6A4) locus in a rigidcompulsive subset of autism.
Am J Med Genet 2004, 127B:104112. Publisher Full Text

Persico AM, Militerni R, Bravaccio C, Schneider C, Melmed R, Conciatori M, Damiani V, Baldi A, Keller F: Lack of association between serotonin transporter gene promoter variants and autistic disorder in two ethnically distinct samples.
Am J Med Genet 2000, 96:123127. PubMed Abstract  Publisher Full Text

Betancur C, Corbex M, Spielewoy C, Philippe A, Laplanche JL, Launay JM, Gillberg C, MourenSimeoni MC, Hamon M, Giros B, NostenBertrand M, Leboyer M: Serotonin transporter gene polymorphisms and hyperserotonemia in autistic disorder.
Mol Psychiatry 2002, 7:6771. PubMed Abstract  Publisher Full Text

Wu S, Guo Y, Jia M, Ruan Y, Shuang M, Liu J, Gong X, Zhang Y, Yang J, Yang X, Zhang D: Lack of evidence for association between the serotonin transporter gene (SLC6A4) polymorphisms and autism in the Chinese trios.
Neurosci Lett 2005, 381:15. PubMed Abstract  Publisher Full Text

Janusonis S: Serotonergic paradoxes of autism replicated in a simple mathematical model.
Med Hypotheses 2005, 64:742750. PubMed Abstract  Publisher Full Text

Walther DJ, Peter JU, Bashammakh S, Hortnagl H, Voits M, Fink H, Bader M: Synthesis of serotonin by a second tryptophan hydroxylase isoform.
Science 2003, 299:76. PubMed Abstract  Publisher Full Text

Gershon MD: Review article: serotonin receptors and transporters  roles in normal and abnormal gastrointestinal motility.
Aliment Pharmacol Ther 2004, 20 Suppl 7:314. PubMed Abstract  Publisher Full Text

Adell A, Celada P, Abellan MT, Artigas F: Origin and functional role of the extracellular serotonin in the midbrain raphe nuclei.
Brain Res Brain Res Rev 2002, 39:154180. PubMed Abstract  Publisher Full Text

Jacobs BL, Azmitia EC: Structure and function of the brain serotonin system.
Physiol Rev 1992, 72:165229. PubMed Abstract  Publisher Full Text

Anderson GM, Stevenson JM, Cohen DJ: Steadystate model for plasma free and platelet serotonin in man.
Life Sci 1987, 41:17771785. PubMed Abstract  Publisher Full Text

Gershon MD: Plasticity in serotonin control mechanisms in the gut.
Curr Opin Pharmacol 2003, 3:600607. PubMed Abstract  Publisher Full Text

Kirchgessner AL, Liu MT, Raymond JR, Gershon MD: Identification of cells that express 5hydroxytryptamine1A receptors in the nervous systems of the bowel and pancreas.
J Comp Neurol 1996, 364:439455. PubMed Abstract  Publisher Full Text

Costall B, Naylor RJ: 5HT3 receptors.
Curr Drug Targets CNS Neurol Disord 2004, 3:2737. PubMed Abstract  Publisher Full Text

Bockaert J, Claeysen S, Compan V, Dumuis A: 5HT4 receptors.
Curr Drug Targets CNS Neurol Disord 2004, 3:3951. PubMed Abstract  Publisher Full Text

Ramamoorthy S, Bauman AL, Moore KR, Han H, YangFeng T, Chang AS, Ganapathy V, Blakely RD: Antidepressant and cocainesensitive human serotonin transporter: molecular cloning, expression, and chromosomal localization.
Proc Natl Acad Sci U S A 1993, 90:25422546. PubMed Abstract  Publisher Full Text  PubMed Central Full Text

Billett EE: Monoamine oxidase (MAO) in human peripheral tissues.
Neurotoxicology 2004, 25:139148. PubMed Abstract  Publisher Full Text

Thomas DP, Vane JR: 5hydroxytryptamine in the circulation of the dog.
Nature 1967, 216:335338. PubMed Abstract

Shiue GG, Choi SR, Fang P, Hou C, Acton PD, Cardi C, Saffer JR, Greenberg JH, Karp JS, Kung HF, Shiue CY: N,Ndimethyl2(2amino4(18)Ffluorophenylthio)benzylamine (4(18)FADAM): an improved PET radioligand for serotonin transporters.
J Nucl Med 2003, 44:18901897. PubMed Abstract  Publisher Full Text

Lin KJ, Ye XX, Yen TC, Wey SP, Tzen KY, Ting G, Hwang JJ: Biodistribution study of [(123)I] ADAM in mice: correlation with whole body autoradiography.
Nucl Med Biol 2002, 29:643650. PubMed Abstract  Publisher Full Text

Cohen IL, Liu X, Schutz C, White BN, Jenkins EC, Brown WT, Holden JJ: Association of autism severity with a monoamine oxidase A functional polymorphism.
Clin Genet 2003, 64:190197. PubMed Abstract  Publisher Full Text

Jones MB, Palmour RM, Zwaigenbaum L, Szatmari P: Modifier effects in autism at the MAOA and DBH loci.
Am J Med Genet 2004, 126B:5865. Publisher Full Text

Filic V, Vladic A, Stefulj J, CicinSain L, Balija M, Sucic Z, Jernej B: Monoamine oxidases A and B gene polymorphisms in migraine patients.
J Neurol Sci 2005, 228:149153. PubMed Abstract  Publisher Full Text

Soeby K, Larsen SA, Olsen L, Rasmussen HB, Werge T: Serotonin transporter: Evolution and impact of polymorphic transcriptional regulation.
Am J Med Genet B Neuropsychiatr Genet 2005. PubMed Abstract  Publisher Full Text

Chen JX, Pan H, Rothman TP, Wade PR, Gershon MD: Guinea pig 5HT transporter: cloning, expression, distribution, and function in intestinal sensory reception.
Am J Physiol 1998, 275:G433G448. PubMed Abstract  Publisher Full Text

Grundy D, Schemann M: Serotonin in the gut: pretty when it gets down to the nitty gritty.
Neurogastroenterol Motil 2004, 16:507509. PubMed Abstract  Publisher Full Text

Cash S, Yuste R: Linear summation of excitatory inputs by CA1 pyramidal neurons.
Neuron 1999, 22:383394. PubMed Abstract  Publisher Full Text

Beck O, Wallen NH, Broijersen A, Larsson PT, Hjemdahl P: On the accurate determination of serotonin in human plasma.
Biochem Biophys Res Commun 1993, 196:260266. PubMed Abstract  Publisher Full Text

Anderson GM, Feibel FC, Cohen DJ: Determination of serotonin in whole blood, plateletrich plasma, plateletpoor plasma and plasma ultrafiltrate.
Life Sci 1987, 40:10631070. PubMed Abstract  Publisher Full Text

Chen JJ, Li Z, Pan H, Murphy DL, Tamir H, Koepsell H, Gershon MD: Maintenance of serotonin in the intestinal mucosa and ganglia of mice that lack the highaffinity serotonin transporter: Abnormal intestinal motility and the expression of cation transporters.
J Neurosci 2001, 21:63486361. PubMed Abstract  Publisher Full Text

Anderson GM, Gutknecht L, Cohen DJ, BraillyTabard S, Cohen JH, Ferrari P, Roubertoux PL, Tordjman S: Serotonin transporter promoter variants in autism: functional effects and relationship to platelet hyperserotonemia.
Mol Psychiatry 2002, 7:831836. PubMed Abstract  Publisher Full Text

McBride PA, Anderson GM, Hertzig ME, Snow ME, Thompson SM, Khait VD, Shapiro T, Cohen DJ: Effects of diagnosis, race, and puberty on platelet serotonin levels in autism and mental retardation.
J Am Acad Child Adolesc Psychiatry 1998, 37:767776. PubMed Abstract  Publisher Full Text

Press WH: Numerical recipes in C: the art of scientific computing. 2nd ed edition. Cambridge, Cambridge University Press; 1992.

Rohatgi VK: Statistical inference. Mineola, N.Y, Dover Publications; 2003.

Chandana SR, Behen ME, Juhasz C, Muzik O, Rothermel RD, Mangner TJ, Chakraborty PK, Chugani HT, Chugani DC: Significance of abnormalities in developmental trajectory and asymmetry of cortical serotonin synthesis in autism.
Int J Dev Neurosci 2005, 23:171182. PubMed Abstract  Publisher Full Text

Chugani DC, Muzik O, Behen M, Rothermel R, Janisse JJ, Lee J, Chugani HT: Developmental changes in brain serotonin synthesis capacity in autistic and nonautistic children.
Ann Neurol 1999, 45:287295. PubMed Abstract  Publisher Full Text

Coon H, Dunn D, Lainhart J, Miller J, Hamil C, Battaglia A, Tancredi R, Leppert MF, Weiss R, McMahon W: Possible association between autism and variants in the brainexpressed tryptophan hydroxylase gene (TPH2).
Am J Med Genet B Neuropsychiatr Genet 2005, 135:4246. PubMed Abstract  Publisher Full Text

WhitakerAzmitia PM: Serotonin and brain development: role in human developmental diseases.
Brain Res Bull 2001, 56:479485. PubMed Abstract  Publisher Full Text

Janusonis S, Gluncic V, Rakic P: Early serotonergic projections to CajalRetzius cells: relevance for cortical development.
J Neurosci 2004, 24:16521659. PubMed Abstract  Publisher Full Text

Hendricks TJ, Fyodorov DV, Wegman LJ, Lelutiu NB, Pehek EA, Yamamoto B, Silver J, Weeber EJ, Sweatt JD, Deneris ES: Pet1 ETS gene plays a critical role in 5HT neuron development and is required for normal anxietylike and aggressive behavior.
Neuron 2003, 37:233247. PubMed Abstract  Publisher Full Text

McKinney J, Knappskog PM, Haavik J: Different properties of the central and peripheral forms of human tryptophan hydroxylase.
J Neurochem 2005, 92:311320. PubMed Abstract  Publisher Full Text

Launay JM, Ferrari P, Haimart M, Bursztejn C, Tabuteau F, Braconnier A, PasquesBondoux D, Luong C, Dreux C: Serotonin metabolism and other biochemical parameters in infantile autism. A controlled study of 22 autistic children.
Neuropsychobiology 1988, 20:111. PubMed Abstract

Croonenberghs J, Verkerk R, Scharpe S, Deboutte D, Maes M: Serotonergic disturbances in autistic disorder: L5hydroxytryptophan administration to autistic youngsters increases the blood concentrations of serotonin in patients but not in controls.
Life Sci 2005, 76:21712183. PubMed Abstract  Publisher Full Text

Kema IP, de Vries EG, Slooff MJ, Biesma B, Muskiet FA: Serotonin, catecholamines, histamine, and their metabolites in urine, platelets, and tumor tissue of patients with carcinoid tumors.
Clin Chem 1994, 40:8695. PubMed Abstract

Siegel GJ: Basic neurochemistry: molecular, cellular, and medical aspects. 6th ed edition. Philadelphia, Lippincott Williams & Wilkins; 1999.

Racke K, Reimann A, Schworer H, Kilbinger H: Regulation of 5HT release from enterochromaffin cells.
Behav Brain Res 1996, 73:8387. PubMed Abstract  Publisher Full Text

Schworer H, Ramadori G: Autoreceptors can modulate 5hydroxytryptamine release from porcine and human small intestine in vitro.
Naunyn Schmiedebergs Arch Pharmacol 1998, 357:548552. PubMed Abstract  Publisher Full Text

Gebauer A, Merger M, Kilbinger H: Modulation by 5HT3 and 5HT4 receptors of the release of 5hydroxytryptamine from the guineapig small intestine.
Naunyn Schmiedebergs Arch Pharmacol 1993, 347:137140. PubMed Abstract  Publisher Full Text

Schafermeyer A, Gratzl M, Rad R, Dossumbekova A, Sachs G, Prinz C: Isolation and receptor profiling of ileal enterochromaffin cells.
Acta Physiol Scand 2004, 182:5362. PubMed Abstract  Publisher Full Text

Wade PR, Tamir H, Kirchgessner AL, Gershon MD: Analysis of the role of 5HT in the enteric nervous system using antiidiotopic antibodies to 5HT receptors.
Am J Physiol 1994, 266:G403G416. PubMed Abstract  Publisher Full Text

Kirchgessner AL, Tamir H, Gershon MD: Identification and stimulation by serotonin of intrinsic sensory neurons of the submucosal plexus of the guinea pig gut: activityinduced expression of Fos immunoreactivity.
J Neurosci 1992, 12:235248. PubMed Abstract  Publisher Full Text

Ahlman H, Larsson I, Lundberg J, Dahlstrom A, Donahue PE, Bhargava HN: Neural control of serotonin release from duodenal enterochromaffin cells of the cat.
Surg Forum 1978, 29:404407. PubMed Abstract

Kuemmerle JF, Kraus H, Kellum JM: Serotonin release is mediated by muscarinic receptors on duodenal mucosal cells.
J Surg Res 1987, 43:139142. PubMed Abstract

Racke K, Schworer H: Nicotinic and muscarinic modulation of 5hydroxytryptamine (5HT) release from porcine and canine small intestine.
Clin Investig 1992, 70:190200. PubMed Abstract

Kahne D, Tudorica A, Borella A, Shapiro L, Johnstone F, Huang W, WhitakerAzmitia PM: Behavioral and magnetic resonance spectroscopic studies in the rat hyperserotonemic model of autism.
Physiol Behav 2002, 75:403410. PubMed Abstract  Publisher Full Text

Varnas K, Halldin C, Pike VW, Hall H: Distribution of 5HT4 receptors in the postmortem human brainan autoradiographic study using [125I]SB 207710.
Eur Neuropsychopharmacol 2003, 13:228234. PubMed Abstract  Publisher Full Text

Vilaro MT, Cortes R, Mengod G: Serotonin 5HT(4) receptors and their mRNAs in rat and guinea pig brain: Distribution and effects of neurotoxic lesions.
J Comp Neurol 2005, 484:418439. PubMed Abstract  Publisher Full Text

McBride PA, Anderson GM, Hertzig ME, Sweeney JA, Kream J, Cohen DJ, Mann JJ: Serotonergic responsivity in male young adults with autistic disorder. Results of a pilot study.
Arch Gen Psychiatry 1989, 46:213221. PubMed Abstract

Goldberg J, Szatmari P, Zwaigenbaum L: Diminished serotonin receptor function: A potential endophenotype of autism.

Chugani DC: Role of altered brain serotonin mechanisms in autism.
Mol Psychiatry 2002, 7 Suppl 2:S16S17. PubMed Abstract  Publisher Full Text

Yu AM, Idle JR, Byrd LG, Krausz KW, Kupfer A, Gonzalez FJ: Regeneration of serotonin from 5methoxytryptamine by polymorphic human CYP2D6.
Pharmacogenetics 2003, 13:173181. PubMed Abstract  Publisher Full Text

Patel S, Dulluc J, Geffard M: Comparison of serotonin and 5methoxytryptamine immunoreactivity in rat raphe nuclei.
Histochemistry 1986, 85:259263. PubMed Abstract  Publisher Full Text

Liu MT, Rayport S, Jiang Y, Murphy DL, Gershon MD: Expression and function of 5HT3 receptors in the enteric neurons of mice lacking the serotonin transporter.
Am J Physiol Gastrointest Liver Physiol 2002, 283:G1398G1411. PubMed Abstract  Publisher Full Text

David SP, Murthy NV, Rabiner EA, Munafo MR, Johnstone EC, Jacob R, Walton RT, Grasby PM: A functional genetic variation of the serotonin (5HT) transporter affects 5HT1A receptor binding in humans.
J Neurosci 2005, 25:25862590. PubMed Abstract  Publisher Full Text

Christofi FL, Kim M, Wunderlich JE, Xue J, Suntres Z, Cardounel A, Javed NH, Yu JG, Grants I, Cooke HJ: Endogenous adenosine differentially modulates 5hydroxytryptamine release from a human enterochromaffin cell model.
Gastroenterology 2004, 127:188202. PubMed Abstract  Publisher Full Text

Bertrand PP: ATP and sensory transduction in the enteric nervous system.
Neuroscientist 2003, 9:243260. PubMed Abstract  Publisher Full Text

Davalos D, Grutzendler J, Yang G, Kim JV, Zuo Y, Jung S, Littman DR, Dustin ML, Gan WB: ATP mediates rapid microglial response to local brain injury in vivo.
Nat Neurosci 2005, 8:752758. PubMed Abstract  Publisher Full Text

Courchesne E, Pierce K: Why the frontal cortex in autism might be talking only to itself: local overconnectivity but longdistance disconnection.
Curr Opin Neurobiol 2005, 15:225230. PubMed Abstract  Publisher Full Text

Vargas DL, Nascimbene C, Krishnan C, Zimmerman AW, Pardo CA: Neuroglial activation and neuroinflammation in the brain of patients with autism.
Ann Neurol 2005, 57:6781. PubMed Abstract  Publisher Full Text